Hot pixel in XL1 CCD block

[James Emory]

I had a shoot last night, 6.6.03, with my XL-1 systems. I suddenly noticed a single dot on the source monitor of the wide shot when the lights went down for the event to start. You can navigate by this dot it so bright when there is a dark background. My first thought was a burned out pixel in the EVF and then the monitor. But the dot is on the monitor and in the EFV. I also noticed that when the gain is turned down the dot also dims and when the gain is turned up it becomes brighter. This could only mean one thing. A bad CCD? Somebody tell me that ain't it, please! Until it's fixed, it shouldn't be a problem for exterior shots but not for anything that has a dark background. Is this a known issue with XL systems? This is one of four identical systems that are now 3 years old and this particular one only has at the most 30 hours on it. The one I carry with me all the time has got to have at least 100 hours or more and it doesn't have this problem. Does anybody know what this is and how it is fixed. Thanks.

[Ken Tanaka]

James,
It sure sounds like a bad pixel on your CCD block. It's not an issue particular to XL1 cameras; it can happen to any digital camera, video or still.

Canon service will be your only remedy.

[Don Palomaki]

Virtually all CCD have hot pixels, they are usually masked electronically. Yours is the first report I' can recall of a visible "hot" pixel on the XL1.

Canon might be able to mask it at the next service. In the interim you maybe able to make a mask to apply in your in your NLE to hide it when necessary.

Specss imply no visible hot pixels at 0 dB gain. If you add gain, the output of a hot pixel can increase above black level and become apparent. This is discussed at the bottom of page 71, About the CCD, in the XL1 manual.

[Chris Hurd]

Canon service will be able to mask it. Although its position may be readily obvious, be sure to include a recording of it on a DV tape with the camera when you send it in.

[Nathan Gifford]

If its more than a single pixel, try taking off the lens filter and see if it goes away. If its several pixels wide you can get some reflections between the lens filter and the lens elements, especially when shooting indoors.

The problem could still (likely) be a hot pixel, but let's elminate the cheap things first.

[James Emory]

Thanks to everyone for a quick response! Unfortunately, I don't think I'll be able to mask this one. It's about 2-3 inches from the top and just a little off center. Just enough to be aggravating. How does Canon mask it. Please excuse my lack of tech knowledge here but from what you all are saying, CCDs are mounted on a single block together. You mean they would not replace the bad CCD? I bet this costs half if not more than the entire system brand new.

Nathan, I sure know what your talking about with re-reflection from the flat and lens especially with chandaliers and spots like headlights. That or a possible dust particle catching a flare was my first thought. But because it was the north star in absolute darkness that was ruled out.

Ken and Don, the fact this is known by the manufacturers and that they release these things to market at these prices is not too comforting. Its like they are playing the odds at our expense.

Hypothetical Manufacturers Logic: "Well, we just hope that a hot pixel won't show, but if it does, the consumer will just have to pay for the repair."

No wonder there is only a 12 month warranty. It's time to call Michael Moore!

[Chris Hurd]

James

The most common technique for masking a hot pixel is to re-map it to an adjacent one (it's easy, cheap and completely transparent). On 3-chip camcorders, it's more cost-effective to replace the entire block than one individual CCD, however, hot pixel masking does not require block replacement, but only an electronic re-assignment of the bad pixel.

I think it's important to look at the cost of current prosumer camcorders with a point of view relative to its own history. $5000 for a 3-chip DV camcorder is *dirt cheap* compared to what you had to pay to get similar quality less than ten years ago. A one-year warranty is a very good warranty by the way, and extended warranties are readily available for not too much money.

[James Emory]

Thanks for the info Chris. I know that for quality that is given, it's a damn good price or I wouldn't have done it. I'm just frustrated right now. It's just bad timing. Do you have any idea what an estimated cost would be for Canon to fix this?

[Chris Hurd]

James, I should know this but I can't recall. If you contact one of the Canon service centers, you might check if they have a base price for this type of work (and please report back if you find out).

[Don Palomaki]

If you research CCD, you will find that hot pixles are a normal sort of thing, even in very expensive CCDs used for astronomy. Most people never see them because they are premasked. But sometimes one will develop later in life, which is why you see one now and did not when the machine was new.

No one could affort a camcorder with a"perfect" CCD where all pixels were very closely matched for sensitivity and dark current.

[Chris Hurd]

<< No one could affort a camcorder with a"perfect" CCD... >>

Don as always is quite correct. One or two hot pixels in a 250,000 pixel CCD equates to a quality rate better than 99.99%, well within acceptable standards, until of course it causes your video to be unacceptable, whereupon you send it in for service and have it mapped out. More expensive camcorders even have a user-accessible menu option for this function.

[James Emory]

Have mercy! It's not as bad as I thought it would be. The Canon repair facility in New Jersey must replace the entire CCD block to fix this burned out pixel. This is only necessary on the XL-1. The XL-1s models can be re-mapped without surgery. Now for the cost. It will cost 500.00 for the new CCD block. The labor will be 400.00 which will include the typical cleaning and factory spec adjustments, if any are needed. And oh yes, 12.00 for shipping. This is major invasive surgery for this camera. They stated that the turnaround time for repairs is around a week.

[James Emory]

Well, after more examination of the footage I found two more bad pixels that were not evident earlier. This is almost like a viral infection. Unbelievable folks!

[Don Palomaki]

Has the camcorcder been subject to any interesting or unusual environmental exposures; e.g., lots of time flying at high altitude?

[James Emory]

3 of the cameras, one of which has the issue, are stored in padded Pelican cases in a climate controlled building. They are spoiled rotten and have very limited hours on them. That's why this is so aggravating. The fourth stays with me and has at least a hundred hours on it and has no issues. The only flying it has seen has been on the jib. I guess this breaks down to 300.00 per pixel. Ha.